PI discrimination is possible

[Teleno]

Coil: Ferrite core 150uH, 01. Ohms,
Pulse duration: 60us
Compensation resistor: 560Ohm.
Vdd = 12V

Oscilloscope:
The pulse in yellow is the start signal of the Tx pulse.
The pulse in blue is the discrimination signal, it shifts right for Fe and left for non-Fe.

LCD screen:
Count is the number of clock pulses (16MHz Arduino) between the yellow and blue signals.

[kt315]

and what? Mustafa does PI's with the disc like cooking hot dogs!
https://www.youtube.com/watch?v=uEXe5N0ZlZE

and all are without the arduino )

[Teleno]

and what? Mustafa does PI's with the disc like cooking hot dogs!
https://www.youtube.com/watch?v=uEXe5N0ZlZE

and all are without the arduino )

Actually, Mustafa's is a piece of **** and you know it.

[ODM]

Seeing actual PI discrimination in varied ground conditions and targets would give more faith besides the usual hidden software tricks or frontend overload.

The issue with PI is general difficulty in processing multiple excitation responses that CW does more elegantly, in MF or sequential sweeping like GEM series does. A detector with a powerful find mode like minelab MPS together with an identify mode with multiple halfsine periods or pulse shapes would likely fare better than a constant excitation pattern with tweaky processing.

The "secret" to discriminating PI won't be in any funny damping circuit or such, since a single excitation pattern won't do it all just as a single frequency won't do it for CW, for analysis purposes. We have seen this in the curious frontends and software tricks so far that have amounted to nothing for target ID outside of known obvious targets in known conditions.

[Teleno]

The "secret" to discriminating PI won't be in any funny damping circuit or such, since a single excitation pattern won't do it all just as a single frequency won't do it for CW

Yes it does, but I'm not gonna tell you the "trick" (hint: it's not a trick, it's Physics).

[moodz]

Yes it does, but I'm not gonna tell you the "trick" (hint: it's not a trick, it's Physics).

I agree with Teleno .. there may also be other ways to do it however "funny" damping and frontend ccts are a winner ... alot harder ( but not impossible ) with standard damping / frontends.

[Teleno]

I agree with Teleno .. there may also be other ways to do it however "funny" damping and frontend ccts are a winner ... alot harder ( but not impossible ) with standard damping / frontends.

My set up is a standard front end. I'm using the turn-on ramp for discrimination instead of the off signal.

This technique can be easily applied to Surf PI, Barracuda etc. as an add-on.

[Qiaozhi]

My set up is a standard front end. I'm using the turn-on ramp for discrimination instead of the off signal.

This technique can be easily applied to Surf PI, Barracuda etc. as an add-on.

Although it is possible to provide some level of discrimination by using the turn-on ramp, it does not provide discrimination for deep targets.

[ODM]

The turn on has a different dI/dt so it is in effect another pulse. There is a patent on some halfsine pulsed and another CC pulsed detector that sample during ontime, during transient. This is more information than a single frequency or single decay detector has, eg. the trick is not in processing decay.

However if there aren't separate ramp profiles or pulse widths, it runs into the same ambiguity problems that single frequency VLF detectors face.

Also on time sampling is not a "trick" or much of a secret, especially with IB coils or separate inductors, please don't lower the standard of discussion on this board back to times of secret "magic formulas and minelab killer detectors".

[kt315]

do in that whites mannere, only delete this 'vdi=' from image
https://www.youtube.com/watch?v=Xr7k28_Pgv0

[Teleno]

Although it is possible to provide some level of discrimination by using the turn-on ramp, it does not provide discrimination for deep targets.

Deep targets are a discrimination problem to any method. Simply dtecting them is already hard enough.

At any rate, discrimination never reaches the same depth as detection. We'll have to live with that.

do in that whites mannere, only delete this 'vdi=' from image
https://www.youtube.com/watch?v=Xr7k28_Pgv0

White's MXT is not PI. FAIL.

[kt315]

Although it is possible to provide some level of discrimination by using the turn-on ramp, it does not provide discrimination for deep targets.

btw Chance of Andy is working in two algorithms - for near targets and for deep targets. much of western people do not know
that.

[ODM]

From my replies to you, Teleno, you know that I do not intend to harm the efforts, only to raise points I believe have not been considered thoroughly, or to point out something done earlier, such as discussing your discrete frontend project.

Your previous postings regarding math in ground cancel are a good example of physics. Physics is models of our world in mathematics, not cloak-and-dagger secrets!

I ask that you don't go down the path other people on this forum have gone and listen to Carl's experience in the industry if you won't listen to me; in previous threads his advice was to either come forward with all information or to keep a lid on a project until it is patentable, if the intention is to make it your own IP. Those are the two ways to go.

Flaunting secret methods and such makes for long threads but ultimately doesn't result in synergy if others cannot inspect process. Scientific method relies on publishing findings so they can be reproduced, but does not discredit the original work!

[Teleno]

The turn on has a different dI/dt so it is in effect another pulse.

Turn-on time depends on inductance. Diamagnetic targets reduce inductance, paramagnetic targets increase inductance. Just simple Physics.

[ODM]

Thinner skin depth will mask target ID based on inductance shift (as in BFOs), so frequency eg. rise time will be critical. Slowing down the LR tc will give thicker skin depth. Some PI engineers suggest series R with coil for other reasons.